Working paper 9

(1)

Working Paper

February 2018

WP 09

AGRICULTURAL

COMMERCIALISATION

PATHWAYS

CLIMATE CHANGE AND AGRICULTURE

Andrew Newsham, Sarah Kohnstamm,

Lars Otto Naess and Joanes Atela

(2)

Contents ... 2

Abbreviations and acronyms ... 5

Summary of key insights ... 6

Introduction ... 7

1 Implications of climate change impacts across commercialisation pathways, and of commercial agriculture for climate change ... 10

1.1 Framing the problem: the relationship between climate change and agricultural commercialisation ... 10

1.2 Global and regional climate impacts: implications for agricultural production ... 11

1.3 Climate impacts, commercialisation and (differentiated) adaptation prospects ... 13

1.4 Gendered impacts of climate change and barriers to commercialisation ... 14

1.5 Implications for livelihood resilience and prospects ... 17

1.6 Climate change impacts on commercialisation from the value chain perspective ... 18

1.7 How resilient to climate change is modern agriculture? ... 19

2 Alternative agricultures for sustainable, climate-resilient commercialisation ... 23

2.1 Conceptualising alternative agricultures ... 23

2.2 Climate-smart agriculture ... 23

2.3 Sustainable intensification ... 24

2.4 The sustainable agri-food system productivity framework ... 26

2.5 Alternative techniques and methods for sustainability in agricultural commercialisation ... 28

2.5.1 Conservation agriculture ... 28

2.5.2 Cover crops ... 29

2.5.3 Agroforestry ... 30

2.5.4 Soil conservation and erosion management ... 30

2.5.5 Irrigation ... 31

ABBREVIATIONS AND

ACRONYMS

(6)

• Climate change is likely to alter the environmental parameters which affect where agricultural commercialisation activity can happen in sub-Saharan Africa, what can be grown and how well it will fare. In other words, it touches upon factors fundamental to the viability of commercialisation pathways.

• Some contemporary forms of agricultural commercialisation – especially industrial agriculture – are an important source of

greenhouse gas emissions and as such contribute to the anthropogenic climate change that farmers subsequently have to adapt to. African agriculture contributes very little to global climate emissions, but the climate impact it will face is contingent, at least in part, on the extent of reform of industrial agriculture elsewhere.

• There is a consensus on high-level findings: temperature will rise, precipitation patterns will change, and cereal crop productivity is therefore expected to decrease. Temperature rise will undermine perennial crop performance, most pests/diseases are likely to increase. A number of areas in which high-value crops including tea, coffee, and cocoa are grown are projected to become less suitable for their production. However, the level of uncertainty that continues to exist in climate projections undermines their credibility as a basis for decision-making in the short-to-medium term even at country level, let alone at farm level.

• Climate change will have an impact on all varieties of commercialisation, but we conjecture that it will differentially affect alternate pathways. There is an insufficient evidence base to substantiate this proposition as yet, and we see an opportunity to contribute to building it through the life course of the Agricultural Policy Research in Africa (APRA) research agenda.

• Vulnerability and resilience to climate impacts is unevenly distributed across people involved in different commercialisation pathways, and can be broken down according to gender, age, ethnicity and class. It is too simplistic to state that women are more vulnerable to climate impacts than men, in relation to agricultural commercialisation; not least because in the sub-Saharan Africa

context, as elsewhere, men are overwhelmingly more involved in commercialisation activities than women. Yet women and men are clearly affected differently by climate impacts. It is therefore key to account for the interactions of these different vulnerability profiles at the household and other levels.

• Whilst agricultural commercialisation activities may be highly sensitive to climate impacts, the vulnerability profiles of the people practising them will depend also on off-farm livelihood activities, which themselves will have their own degree of climate sensitivity (i.e. manual labour in warmer conditions).

• There are differing views over whether the modernising agricultural technologies often associated with agricultural commercialisation increase or reduce vulnerability to climate impacts. The answer to this question has fundamental ramifications for the kinds of commercialisation pathways it will be advisable to invest in if questions of socio-environmental sustainability are to be addressed robustly, along with those of gender empowerment and poverty reduction. Correspondingly, across commercialisation pathways, a key consideration will be around the extent to which changes to farming strategies currently envisaged or taken by those seeking to commercialise, or by those seeking to maintain existing commercialisation activities, strengthen or weaken resilience in the face of climate impacts. • Policy approaches that aim to foster agricultural

commercialisation are not automatically reconcilable with efforts to enhance adaptive capacity in the face of climate impacts. Nevertheless, there are agricultural practices that offer the prospect of increasing resilience to climate impacts and productivity, and these are documented in this paper. The extent to which any of them, or the conceptualisations in which they are based, such as sustainable intensification or climate-smart agriculture, offer prospects for sustainable agricultural commercialisation pathways at scale, is contested. However, using an approach such as Sitko and Jayne’s (2017) ‘sustainable agri-food system productivity’ framework, we can get a clearer sense of which practice or technique is more or less likely to work in specific commercialisation contexts.

(7)

1. INTRODUCTION

Rationale and aims

This paper presents a review of recent literature on the implications of climate change for agricultural commercialisation, focusing chiefly on sub-Saharan Africa, and incorporating evidence, where relevant, from around the world. Climate change is one of the crosscutting themes of the Department for International Development (DFID)-funded Agricultural Policy Research in Africa (APRA) consortium.1 _APRA is intended to produce new data and insights into agricultural commercialisation processes, and their impacts and outcomes with regard to rural poverty, empowerment of women and girls, and food and nutrition security. In addition to outlining our rationale and aims, this introduction sets out (a) the approach we have taken to classifying climate impacts upon agricultural commercialisation, and (b) the structure. Time-constrained readers wishing quickly to get a sense of the paper’s principal insights and recommendations are directed to the summary on page 6.

Given the highly climate-sensitive character of agricultural production, climate change has obvious and important ramifications for agricultural commercialisation, which in turn have a bearing on poverty, gender empowerment, and food and nutrition security. The nature and extent of climate change implications for agricultural commercialisation will depend on the choices that are made and the resulting commercialisation pathways. It is therefore important that there are opportunities to build considerations of climate change into the design of the APRA research in the early stages. With this objective in mind, this paper aims not only to give a broad indication of the current state of knowledge on climate change and agricultural commercialisation, but also to indicate how and where this is relevant to the APRA research priorities, and to identify key questions to explore in its data collection activities. As such, we outline the salience of key issues for each of the three work streams that, broadly, comprise the APRA research effort. Additionally, we offer a preliminary exploration of the overlaps and intersections between

the different crosscutting research themes.

Types of climate impacts upon

agricultural commercialisation

The implications of climate change for agricultural commercialisation can, for the purposes of analysis, be grouped into two related sets. The first set might be called agro-biophysical: climate change is projected to alter the environmental parameters within which agricultural activity takes place. Higher temperatures, in combination with changes in rainfall quantities and patterns across Africa, may have profound effects on what can grow and where, with potentially fundamental consequences for the types and distributions of cultivation and livestock farming, and even for the viability of these activities. To be sure, the exercise of attempting to predict how climate change will unfold, let alone how it will affect agricultural activities, is fraught with deep uncertainty. It is increasingly clear that climate change projections are not sufficient to form the basis of decision-making about responding to climate change impacts, at least in the short-to-medium term. It is for this reason important to clarify caveats around interpreting and using such information. Yet commercialisation hinges on the most marketable crops, and it is, therefore, important to consider even at a general level the magnitude of the impacts that climate change is currently envisaged to have for some of these. Thus rather than planning for specific climate change scenarios, the focus should be on planning for robustness in the face of a range of possible future climates, and gaining a clear sense as possible of existing adaptive capacity and vulnerability dynamics in the face of historical and contemporary forms of climate variability.

The second set of implications is societal, understood as an umbrella term incorporating social, political and economic dimensions. It is crucial to attempt to understand the implications for vulnerability and resilience to climate change and variability, differentiated chiefly by gender, age, class and ethnicity, for those involved in agricultural commercialisation

(8)

activities and pathways. Climate impacts have different consequences for women and men, boys and girls, because women are involved in, or excluded from, commercialisation activities in sometimes strikingly different ways. Yet vulnerability and resilience also need to be understood relationally, both between women and men, and between different social groups. Only in this way can we start to grasp the implications of climate impacts for rural poverty, empowerment of women and girls, as well as food security and nutrition.

However, once these different impacts have been identified, it is then important to examine the interactions between them. Therefore, we seek to understand and outline not only the implications of climate change for agricultural commercialisation, but the implications of agricultural commercialisation for climate change itself. These latter implications potentially change the nature of what we have to adapt to, how able we are to do that, and who is most adversely affected. It is well known that agricultural activity (both commercial and subsistence) contributes substantially to greenhouse gas emissions, but there is an increasing literature which suggests that some forms of commercialisation also weaken social-ecological resilience to climate impacts, even as they offer benefits for poverty reduction and food security. Again, the idea of relational vulnerability is helpful in understanding these dynamics. Additionally, debates around transformation, which are becoming increasingly prevalent not just in relation to climate change but across a broader range of sustainable development issues, are also salient. Agricultural commercialisation is often seen as part of a broader process of structural

transformation in a country’s economic development. Historically speaking, it could be argued, the processes of structural transformation we have witnessed have given rise to forms of large-scale, intensive commercial agriculture which beg rather than answer the climate question. If the future is not to repeat the past, then debates around transformative climate adaptation

may yield insights into directions and forms that commercialisation activities might seek to take.

Structure of this paper

This paper is divided into three main sections. There is a strong focus in the structure towards serving the objectives of the APRA consortium. This is particularly the case with Section 4, but Sections 1–3 do not require prior knowledge of the consortium’s aims. Section 1 has two main aims. First, it introduces the relationship between global agricultural production and climate change, outlining the central conundrum facing those involved in agricultural commercialisation: agriculture is both a key driver of climate change and highly vulnerable to climate impacts. Second, it surveys the implications of climate change for agricultural commercialisation activities across sub-Saharan Africa, both in terms of (a) the implications of climate impacts for the viability of agricultural production and associated value chains, and (b) the implications, gendered and otherwise, for livelihoods resilience. Section 2 considers the extent to which the modern methods of agricultural production which most obviously suggest themselves for commercialisation are climate resilient. This paves the way for a consideration of alternative concepts, methods and techniques for sustainable, resilient commercial agriculture. At this juncture we introduce Sitko and Jayne’s (2017) ‘sustainable agri-food system productivity’ framework, which is useful for assessing the prospects for these different practices and techniques.

All of the issues that we cover in Sections 1 and 2 are relevant across agricultural commercialisation pathways. In Section 3, we examine the implications of climate change which we take to be specific to the particular commercialisation pathways at the heart of the APRA research agenda (see Box 1). Each pathway is characterised by differences in factors such as technologies deployed, farming systems, crops Box 1: Four pathways of commercial farming

1. Estate/plantation/large-scale commercial farming – large landholding, typically growing a single cash crop, high level of mechanisation, relies on labour (Smalley 2013).

2. Outgrowing/contract farming – farmers supply produce to a central buyer on a contractual basis (see Oya 2012 for a discussion on various contract farming business models).

3. Medium-scale enterprises – more than 5–100 hectares, often owned by urban-based investors who hire labour to produce (Jayne et al. 2014; Sitko and Jayne 2014).

4. Smallholder farmers – rural, farmed by household members, full-time producers who sell surplus/cash crop.

(9)

grown, gendered divisions of labour, and access to productive resources. We therefore conjecture that the implications for climate change for different agricultural commercialisation pathways will differ precisely because of the divergent, specific characteristics of each pathway. This is a necessarily hypothetical statement: there is not yet a sufficiently strong evidence base on climate impacts on specific commercialisation pathways which would allow us to make definitive statements. We therefore present and reflect on the existing evidence, but suggest also that there is an opportunity for APRA to contribute to building knowledge in this area. Section 4 is devoted to teasing out the ways in which the three work streams that comprise APRA can incorporate climate considerations, and conduct a similar exercise for the crosscutting research themes.

(10)

1. SCIENCE, TECHNOLOGY AND

INNOVATION IN AFRICA’S DEVELOPMENT

1.1 Framing the problem: the

relationship between climate change

and agricultural commercialisation

From the perspective of commentators concerned with climate change, agricultural commercialisation is a conundrum. This is in essence because it is implicated at a fundamental causal level in generating anthropogenic climate change, and at the same time one of the modes of economic activity most sensitive to climate impacts.

Most commonly, the greenhouse gas emissions from agriculture, as a percentage of the total global anthropogenic emissions, are accounted for along with forestry and other land uses. On this measure, they contribute to 24 percent of global greenhouse gas emissions – see Figure 1. According to the Intergovernmental Panel on Climate Change (IPCC) (Smith et al. 2014), agriculture is the largest single contributor to global anthropogenic non-CO₂ greenhouse gas emissions, accounting in 2005 for 56 percent of such emissions. In 2010, annual non-CO₂ greenhouse gas emissions were estimated at

the equivalent of 5.2 5.8 gigatonnes of CO₂, or 10–12 percent of global anthropogenic emissions (FAOSTAT 2013). Smith et al. (2014) report that there is not full agreement between the main databases used to disaggregate the proportion of greenhouse gases generated by different agricultural activities. However, there is agreement that enteric fermentation2_and agricultural soils represent together about 70 percent of total emissions, followed by paddy rice cultivation (9–11 percent), biomass burning (6–12 percent) and manure management (7–8 percent). Some appraisals have gone further than agriculture itself, and have sought to estimate the contribution of the global food system – including elements such as indirect emissions associated with land cover change, fertiliser manufacturing, food storage and packaging – to anthropogenic emissions. Vermeulen et al. (2012) put that figure at 19–29 percent of anthropogenic emissions, based on 2008 data. It is possible to extrapolate this figure further, if the emissions generated not only through the production of food but the carbon intensity of international trade in food are taken into account.

Broadly speaking, the more commercial the agriculture, the more industrial the mode of agricultural production tends to be. The more industrial the mode of production, the greater its contribution to the greenhouse gases driving the biggest global environmental problem we face. At the same time, the vulnerability of agriculture and food systems to disturbance (via climate impacts or otherwise) has increased (Beroya-Eitner 2016). In other words, one of the biggest threats to the prospects for commercial agriculture is commercial agriculture itself, at least as currently practised. This dynamic is fundamental to any discussion of commercialisation pathways, in sub-Saharan Africa or anywhere else. This is essentially why climate change has been chosen as one of the crosscutting themes for the APRA project. If agricultural commercialisation is to remain a viable proposition in the medium-to-long term, it would appear that its modes of production need to change fundamentally. However, it needs to be recognised that not all forms of commercialisation have contributed equally to climate change: Africa’s contribution to

Figure 1: Global greenhouse gas emissions

by sector

Electricity and heat production

25%

Agriculture , forestry and other land use

24% Transportation 14% Industry 21% Other energy 10% Buildings 6% Source: IPCC (2014).

(11)

agriculture-related emissions – or indeed to greenhouse gas emissions more broadly – remains minimal. The industrialisation of agriculture in countries like the United States, and the widespread use of industrial methods of intensive agriculture, are far more important contributors than, for instance, increased maize yields produced by Zimbabwean farmers who were given land in the country’s land reform process in the 2000s. Ultimately, some commentators argue that a thoroughgoing solution to this conundrum requires a process of agricultural de-intensification in the places in which it is most industrially practised, even whilst sustainable intensification could be permitted in regions which do not currently contribute very much to global greenhouse gas emissions (Struik and Kuyper 2017). Getting smallholder farmers in sub-Saharan Africa with an interest in reaching a bigger market to adopt conservation agriculture is not going to make much of a dent in global climate emissions attributable to the agricultural sector. There are also powerful ethical considerations relating to the historical and contemporary responsibility for greenhouse gas emissions of different parts of the world. Moreover, sustainable intensification of commercialisation pathways in Africa will remain a hostage to fortune of the broader tendency of global industrial agriculture to contribute to the conditions for a more hostile future climate. However, if commercialisation in Africa, with a view to empowering women and girls, is to be commercially and environmentally viable over the medium-to-long term, then sustainable intensification may offer better prospects than conventional agricultural industrialisation. Furthermore, given the importance of at least maintaining, and ideally increasing, crop yield per level of input as a prerequisite for commercial activity – not to mention local, regional or global food security – it would seem necessary to explore the options offered by and the prospects for sustainable intensification. In view of this discussion, this section therefore has two aims. First it discusses the known implications of a changing climate on agricultural production, partly in terms of its agro-ecological implications but also in terms of other important dimensions, including value chains and social impacts, such as those that are gendered in character. Second, it outlines some of the practices and techniques most associated with the concept of sustainable intensification. This section thereby informs the later sections on the implications of climate change for specific commercialisation pathways: large-scale commercial farming, outgrowing, medium-large-scale commercial farming, and commercialising smallholder farmers.

1.2 Global and regional climate

impacts: implications for agricultural

production

Climate change is, it is argued increasingly, already changing agriculture. There is increasing evidence of climate impacts on production, and climate change is significantly affecting priorities and actions in the agricultural sector, notably through investments in ‘climate-smart agriculture’. The IPCC (2014) reports that according to global, regional, national and local studies across countries and crops, the effects of climate change have been more negative than positive to date. Impacts are, however, variable. Climate change has led to both increases and decreases in agricultural yields, depending on the change in precipitation and temperature and its positive or negative effect on the crop planted (ibid.: 491). It is reported to have reduced crop yields overall, with notable reduction in the global production of wheat and maize as well as small negative effects on global rice and soybean production. Conversely, there has also been a net benefit to ‘crop production in some high-latitude regions, such as northeast China or the UK’ (ibid.: 491). According to the IPCC, climate change affects agricultural crop production by changing rainfall, increasing frost damage, and increasing temperatures (ibid.: 492–93). For another summary of the relevant global scenario studies on climate change impacts, global environmental change, changes in ecosystem services, and changes in agriculture, see also van Vuren et al. (2009).

Looking to the future, researchers using climate scenarios project that there will be ‘increases in air and soil temperatures, changes in CO2 concentration in the atmosphere, sea level rise, changes in the hydrological cycle and in water quality and availability, intensification and increase in frequency of extreme weather, including droughts and floods, changes in the altitudinal level of dew points’, amongst other impacts with fundamental implications for agricultural activity (Vergara et al. 2014: 1). According to a 2011 global review,3_{the changes in} the water cycle will include increased floods, mudslides, river erosion, storms, drought, sea level rise, changes to the length and stability of rain throughout the growing season, shifts in pest and disease problems, and increased evaporation from plants and soils (Toulmin 2011). These environmental changes are expected to impinge upon agricultural production systems worldwide.

In Africa, the IPCC reports that climate change will result in negative effects on yields of major cereal

(12)

crops, including maize at lower elevations (where the majority of current maize production occurs), wheat, and to a lesser extent sorghum (IPCC 2014: 1218). While some areas of East Africa and central Africa may become optimally suitable for maize production, today’s ‘maize-based systems, particularly in Southern Africa are among the most vulnerable to climate change’ with estimated losses of 18 percent by 2050, 22 percent across sub-Saharan Africa (ibid.: 1218). Simulations anticipate high vulnerability in wheat production with negative effects of 35 percent by 2050 (ibid.: 1218). In West Africa, precipitation is projected to increase but potential benefits are likely to be offset by temperature rise, entailing a lower likelihood of positive effects for millet or sorghum yields. Negative effects are likelier to be seen in the savannah. Of particular relevance to the prospects for agricultural commercialisation, negative effects are projected to be higher ‘with modern cereal varieties compared with traditional ones’ (ibid.: 1218). Impacts on crops are anticipated to be particularly negative for the region because they are already grown ‘close to their limits of thermal tolerance (Conway 2009, in Phiiri et al. 2016: 59).4

Overall, in the absence of appropriate adaptation, crop yields in some African countries could decline by as much as 50 percent by 2020, under particular scenarios (IPCC 2007). Yield declines of this order of magnitude would put greater pressure on food access, due to increases in prices of major food crops such as wheat, rice and maize (Nelson et al. 2010).

The IPCC reports that projections for implications of climate change indicate variable impacts on non-cereal crops, with losses in different areas for different

crops and changes in suitable growing areas in Africa. For example, cassava yields could increase in eastern and central Africa. Cassava is hardy enough to withstand high temperatures and sporadic rainfall, and in consequence ‘it may provide a potential option for substitution of cereals as an adaptation response to climate change’ (IPCC 2014: 1218–19). While banana and plantain production may decline in West and lowland East Africa, it may increase in the East Africa highlands (ibid.: 1219).

Perennial crops may face serious challenges, with projected high losses in the production of high-value crops including tea, coffee, and cocoa. The areas that are currently suitable for these crops are projected to become less so (see Table 1). Because these crops take years to come into full production, requiring resource investment up front, and because many perennial crops are grown under contract farming arrangements, yield reduction or failure is potentially a serious problem for producers in particular areas. Conversely, it is a potential opportunity for producers in other areas. Mapping potential changes in these areas, to indicate the implications for the outcomes of commercialisation, and the winners and losers from these changes, could be relevant to the aims of all three APRA work streams, but would ultimately depend upon the level of certainty around projections of change in underlying growing conditions.

There is a large body of literature on the likely direct impacts of climate change on agricultural output and crop production impacts at regional levels.5_In 2013, the International Food Policy Research Institute (IFPRI) published a series of comprehensive regional

Table 1: Projected changes in agro-climatic suitability for perennial crops in Africa by

mid-century under an A2 scenario

Crop Suitability change Country Source

Coffee Increased suitability at high latitudes Decreased suitability at low latitudes

Kenya Läderach et al. (2011)

Tea Decreased suitability Uganda Eitzinger et al. (2011a,

2011b) Increased suitability at high latitudes;

decreased suitability at low latitudes

Kenya Cocoa Constant or increased suitability at

high latitudes; decreased suitability at low latitudes

Ghana, Côte d’Ivoire Läderach et al. (2011)

Cashew Increased suitability Ghana, Côte d’Ivoire Läderach et al. (2011) Cotton Decreased suitability Ghana, Côte d’Ivoire Läderach et al. (2011) Source: IPCC (2014: 1219).

(13)

publications for East, West and South Africa that provide country-level analysis of climate projections, The Climate Change in Africa series. The relevant country data is given in Annex 1 of this paper. It is, however, necessary to clarify that the information offered by climate projections is inevitably uncertain, given data and modelling constraints. For instance, when Namibia first attempted to model the impacts of climate change upon agriculture at the country level, using downscaled regional models (Dirkx et al. 2008), three of the scenarios suggested an increase in rain, whilst three suggested a decrease, in the areas of the country most intensely cultivated. Adapting to a decrease in rainfall would require a very different set of interventions to adaptation to an increase in rainfall, and the information provided in these projections was not a sufficient basis for making policy decisions about agricultural adaptation (Newsham and Thomas 2009). Furthermore, the annual average rainfall projections offered by models did not pick up on rainfall variation experienced within a given growing season, even though this can be fundamental to the size of crop yields. For instance, farmers practising rainfed agriculture in North-Central Namibia reported changes from one growing season to the next in the arrival and intensity of rain, with significantly adverse implications for maize and pearl millet crops (Newsham and Thomas 2011). In a more recent review of climate change in West Africa, Bamba Sylla et al. (2016) confirm that temperature increased in recent decades and projections indicate stronger warming in the future. The authors present evidence that although the Sahel has recovered from previous droughts in the 1970s and 1980s, projections show a significant decrease in rain for all of West Africa in the future, most markedly in the westernmost Sahel. Overall, the region is expected to experience a shorter growing season, shorter rainy seasons, extended torrid, arid and semi-arid climate conditions, longer dry spells and more intense extreme rain, all of which are expected to negatively affect agricultural production (ibid.). In the same book, Pardo et al. (2016) confirm that projections in rainfall patterns for the region generally agree that inconsistency will increase, resulting in greater seasonal variability – but projections cannot yet clarify the direction of change, with some predicting an increase while others predict a decrease (ibid.). While detailed and localised projections remain unclear, climate in the region is unmistakeably changing; Yaro et al. (2016) present the case of Northern Ghana where farmers are adopting adaptation strategies to deal with the climate stress they are experiencing, such as irrigation, use of fertilisers, and modern agronomy, shifting to non-farm activities, and migrating seasonally.

Even more uncertain is our understanding of the changing character of future human agricultural activity, and its implications either for mitigation or adaptation (Ensor 2009; Lipper et al. 2014). Ensor argues that ‘optimising agricultural, aquacultural or pastoral strategies to a particular climate future risks maladaptation and will always be a mistake as long as uncertainty remains in climate projections’ (2009: 6). The projections on which country level analysis offered in Annex 1, and especially the extent to which they can serve as a basis for decision-making, should be considered against this background. The summary intentionally highlights whether or not the various models used were in agreement in their findings, which may indicate a more plausible anticipated scenario, or whether they were not in agreement, and are therefore unable to indicate potential implications.

1.3 Climate impacts, commercialisation

and (differentiated) adaptation

prospects

The geography of the physical implications of climate change is becoming clearer, and what is already known is extremely relevant for researchers, policymakers, and farmers to understand the implications of climate change for different types of farming in different places. Suitable growing areas will shift, and the winners and losers will change. But in order for anyone to benefit from these shifts, farmers will need to anticipate the changes and adapt to grow the right crops in the right areas at the right time, anticipating and adapting to climatic changes that may differ from one season to the next. There is therefore a related large body of literature on climate information dissemination (Kadi et al. 2011a, 2011b; Policarpio and Sheinkman 2015). Men are more likely to be connected to sources of climate information than women (Skinner 2011), as are organised farmer groups and better-off farmers. For example, Thomas et al. (2007) investigate farmer response to rainfall variability in three regions in South Africa to see whose coping and adaptation is stronger; two of the three had stronger adaptation due to collective action in farming cooperatives where they were able to pool resources, information, and advice. There is also evidence that men have more access than women to information on climate-smart agriculture and agricultural extension (Jost et al. 2015; Huyer et al. 2015).

Likewise, incentives are likely to change. For example, the private sector may invest in contract farming schemes for coffee or cocoa in new or anticipated suitability areas or increasing risks and uncertainty may

(14)

discourage companies from investing. Smallholders, medium-sized farmers, and estate farmers are likely to face different opportunities and risks depending on the changes in the natural environment, as well as policy environment as governments, private sector and citizens navigate changing suitability of areas for major crops.

Women may face particular barriers. Carr and Thompson (2014) cite 11 studies that demonstrate that men and women grow different crops, resulting in different vulnerabilities to climate change. They report that this body of literature concludes that ‘women often raise crops that are more sensitive to climate variability than men’ (ibid.: 184). Conversely, whilst men may grow hardier crops on average, to the extent that they are likelier to be growing commercial crops, and to the extent that these are more sensitive to climate impacts than better adapted local crops with lower commercial potential, men’s vulnerability profiles may be more relevant to the consideration of agricultural commercialisation prospects. Understanding both individual male and female vulnerability profiles, and in relation to each other, through social relations such as household units, class and ethnicity, is key to getting a handle on adaptation and commercialisation dynamics. There are several large bodies of literature related to how to manage the implications for agriculture, which are relevant to this review. There is a large body of literature focused on related technical developments and opportunities, including biotechnology, seed varieties – both new and traditional – that are drought- and flood-resistant as well as the implications of climate change for commercialised agriculture by crop type and geographic region (Hertel et al. 2010; La Rovere et al. 2010). 6

1.4 Gendered impacts of

climate change and barriers to

commercialisation

A brief review of the literature on the gendered impacts

of climate change reveals a changing narrative that is important to understanding gender-related barriers to commercialisation. A large literature exists on women being more vulnerable to and disproportionately affected by climate change (e.g. AFDB 2011; CARE 2010; Skinner 2011). However, a new wave of literature now calls for an approach to gender in climate change that goes beyond generalisations about women’s vulnerability, acknowledging the intersections between gender, caste, ethnicity, age and other important drivers of exclusion and vulnerability (Arora-Jonsson 2011, 2014; Jost et al. 2015; Gonda 2016; Twyman and Ashby 2015). Table 2 provides an overview of some commonly noted gendered vulnerabilities and the relationship with climate change.

Gender roles, gendered access to resources, and gender-constricted access to power may contribute to climate vulnerability in specific contexts. However, reappraisals of gender and vulnerability to climate impacts argue that they may not be generalised. According to Jost et al.: ‘A main challenge for the climate change research community is to move beyond the current simplistic understanding of smallholder women as a homogenous group that is inherently nature-protecting, but unable to adapt to climate change because of their overwhelming vulnerability’ (2016: 142). Indeed, narratives that cast women and girls in the light of victimhood and vulnerability may have important implications for female access to opportunities for commercialisation and for the kinds of adaptation policies that are aimed at them.

In March 2015, CGIAR’s research programme on Climate Change, Agriculture and Food Security (CCAFS) hosted a workshop, Closing the Gender Gap in Farming Under Climate Change. The keynote speakers highlighted common myths that had permeated the climate change community about women, namely that they are more vulnerable to climate change because they are poorer, they make better stewards of natural resources than men, and new climate-smart technologies can close the gender gap. But they presented evidence that men and women do have different vulnerabilities to climate change, which are related to gender norms, division of labour, access to and control over resources, and decision-making power, which coincides with earlier literature. They presented the argument that these issues must be placed in context, referring to the case made by Arora-Jonsson that focusing on women’s vulnerability or virtuousness can deflect attention from inequalities and have negative effects (Twyman and Ashby 2015).

Box 2: Increased commercialisation in Africa • Population growth is driving demand;

• Connectivity between rural and urban centres is improving;

• Grocery outlets and retail opportunities are growing.

(15)

Arora-Jonsson (2014) provides an example of a climate programme that intended to influence gender norms by granting land tenure to women, but the programme did so regardless of the womens’ preference or intention. She describes the programme in Andhra Pradesh in India that awarded individual land title and wages through a government scheme (NREGA) in return for planting cash crops on their land:

These crops made the women vulnerable to irrigation problems, volatile markets, and agricultural chemicals and resulted in a total change in their livelihoods. In some districts women were ‘encouraged’ to grow biodiesel plants (Pongamia pinnata) as part of climate programs that would enable them to earn carbon credits. The degraded forests, which they would have regenerated with indigenous species, and agriculture lands that supported food crops were replaced with mono-plantations as they were assured a regular income from the sale of seeds. After one payment from the World Bank for

neutralising carbon emissions, a few years down the line, 80 percent of the trees perished, most families were forced to sell their cattle, were subject to an increased dependence on chemicals and ruined their land in the process. What also emerged was that the women were completely unaware of the reason they had received the money and had no idea about the ramifications of carbon trade and the relationship of their self-help group activities to climate change (Ramdas 2009, in Arora-Jonsson 2014: 301–02)

Arora-Jonsson (2014) argues that the focus on women in climate change programmes can curb agency when it is not part of a gendered analysis of neoliberal policies, technocratic solutions, and decision-making related to climate change that is often performed in high-level meetings with governments, international non-governmental organisations (INGOs), companies and scientists, and without the participation of the women who are intended to ‘benefit’ and/or carry out the work of the programmes.

Table 2: Gender-differentiated vulnerability to climate impacts

Women Men Link to climate change vulnerability

Stay home to care for children, as well as sick or elderly family members

Can migrate to access economic opportunities

Their ability to migrate in search of economic opportunities makes it easier for men to deal with crisis, and may result in benefits for the family as a whole. However, male migration often increases women’s workload, as they are left behind to manage the household in addition to usual tasks. It can also increase women’s exposure to other risks, such as gender-based violence and HIV infection.

Produce household-oriented crops and livestock products

Likelier to produce market-oriented crops and livestock products

Both crops and livestock are affected by climate change, and this has profound consequences for household food security. Men often claim safer/ more fertile land for growing market-oriented crops, leaving women to grow household-oriented crops on more vulnerable/less fertile land. Are responsible for food storage

and preparation

Are responsible for selling valuable produce and livestock

In addition to the challenges described above, climate change has implications for food preparation and storage (in terms of water for food preparation and the vulnerability of food stores to extreme events, such as cyclones and floods). Harvests may be reduced or even wiped out by floods or droughts. This affects market prices and the availability of surplus to sell – placing pressure on both men and women to identify other sources of income and reduce major expenditures (e.g. school fees). In times of food shortage, women are often expected to feed other members.

(16)

Gonda (2016) provides another example through her research in Nicaragua on a climate adaptation project. She finds that the project was based on generalised narratives about women’s roles and vulnerabilities, and thereby reinforced gender stereotypes and roles and was out of sync with reality. In her study, she found that men and women typically shared household duties of collecting water and firewood in the community, and it was the non-governmental organisation (NGO) workers’ assumptions that introduced ‘improved’ cooking stoves and water reservoirs intended for women’s benefit, but that actually benefited both genders (ibid.). Efforts to ensure that agricultural commercialisation contributed to women and girls’ empowerment would do well to avoid similarly imprecise generalisations.

Carr and Thompson provide a review of gender and climate change adaptations in agrarian settings, finding that ‘the contemporary literature on adaptation widely acknowledges that the patterns of vulnerability to climate change impacts we see today are largely, if not principally, shaped by roles, responsibilities

and entitlements associated with various markers of social status and expectation including gender, class and caste’ (2014: 183). However, they argue that the literature on gender and adaptation ‘makes its case through very narrow binary gender analyses, where “man” and “woman” are treated as unitary categories with contrasting needs’ without adequately situating the gender analysis within ‘a host of other social markers like age, income and ethnicity’ (ibid.: 183). For example, they argue that the vulnerability of a wealthy woman’s livelihood may have more in common with a wealthy man, than with a poor woman’s; the majority of their excellent discussion is related to the small but growing literature that deals with adaptation at this level of analysis.

Another example of this change in the literature can be seen in the discussion on gender roles. There is evidence that gender roles contribute to women’s greater vulnerability to climate change. Romero González et al. (2011) found that in Burkina Faso, despite the negative implications of climate change for housework as well as agricultural tasks, women maintain their reproductive Have lower incomes and are

more likely to be economically dependent

Have higher incomes, likelier to own land/other assets

Men typically have more money and other assets than women. Men’s savings provide a ‘buffer’ during tough times and, along with other assets, make it easier for them to invest in alternative livelihoods.

Have less access to education and information

Have more access to education and information

Managing climate-related risks to agricultural production requires new information, skills and technologies, such as seasonal forecasts, risk analysis and water-saving agricultural practices. Men are more likely to have access to these resources and the power to use them and are, therefore, better equipped to adapt. At the same time, women often have traditional knowledge that can inform adaptation efforts. Both new and old information is important in the context of adaptation.

Have less power over family finances and other assets

Have more power over family finances and other assets

Without the power to decide on family resources and finances, women’s ability to manage risks by, for example, diversifying crops, storing harvested crops, etc. may be reduced.

Have limited engagement in community politics

Have greater decision-making power in community politics

Men are likely to have more influence over local governance-promoting policies and programmes that may not support women’s rights and priorities. Face many cultural restrictions

on mobility

Face few cultural restrictions on mobility

Mobility is a key factor in accessing information and services. It is also critical for escaping the danger posed by extreme weather such as floods. Therefore, women are often at higher risk from these events.

Source: CARE (2010). Available online: www.careclimatechange.org/files/adaptation/CARE_Gender_Brief_Oct 2010.pdf

Roles

(17)

and care roles while taking on more productive roles with no re-distribution of access to and control over assets, making them more vulnerable to climate change than men.

However, Okali and Naess (2013) argue that while gender roles may contribute to women’s vulnerability to climate change, it is essential to acknowledge social complexity beyond arguments that essentialise women’s and men’s roles in order to understand the complex and layered barriers to women’s full inclusion. For example, while women farmers may have limited access to inputs, even when they have equal access, they produce less than their male counterparts because of other barriers to inclusion. A recent study by the World Bank and the ONE Campaign (2014) argues that the drivers of the gender gap in agriculture are beyond the focus on access and gender roles, including less household labour availability, less hired labour, lower use of and return on inputs, less effective extension services, less access to land and markets, less human capital and access to social networks, and poorer soil quality. The study provides country-specific analysis of the gender gap for agriculture in Ethiopia, Malawi, Niger, Nigeria, Tanzania and Uganda.

Women face climate change impacts in the context of barriers to commercialisation as well.7_{According to a} recent review, these include: lagging land ownership, less decision-making power within and outside the household, lower levels of access to and control over key agricultural resources namely finance and credit, less access to agricultural inputs, and less access to extension services, technology, training and information including climate information (Huyer et al. 2016). Some commentators go even further than identifying specific barriers to entry in commercialised agriculture for women. For instance, both Akram-Lodhi (2016) and O’Laughlin (2009) contend that women are not involved in commercialisation activities precisely because these are governed by a patriarchal division of labour in which men seek either to profit from their own labour or to sell it, whilst it falls to women to occupy themselves with the unpaid work associated with reproducing labour. It is possible that this underlying gendered division of labour has implications as adverse, and possibly more so, for the involvement of women and girls in commercialisation activities, and their likelihood of benefiting from them, than climate impacts.

Dancer and Tsikata (2015) review gender and commercialisation in various forms, small- and medium-scale, contract farming, and estate/plantation farming, in Africa in order to look at sources of women’s

subordination in the commercialisation process. They find key barriers to women’s advancements in commercialisation to be: (a) lack of land ownership and title, land loss and displacement, (b) labour patterns and work burdens, including the impacts of mechanisation, and (c) constraints in bargaining power and control over resources (ibid.). While their excellent review of the literature sheds light on how commercialising pathways are shaping women’s access to resources and power relations in Africa, it does not consider implications of climate change on these pathways. This review attempts to look at the literature on implications of climate change on these pathways in order to better understand what the implications may be for the women, men, girls and boys engaged in agricultural commercialisation activities along the various pathways.

1.5 Implications for livelihood

resilience and prospects

While there is large literature on the implications of the direct physical implications to agricultural commercialisation as outlined in the first section of this review, there is less evidence on the implications of climate change for women, men, girls’ and boys’ livelihood prospects in agricultural commercialisation activities.

Commercialisation of agriculture is uneven, with women facing more barriers than men (Dancer and Tsikata 2015); therefore, both groups face climatic changes and shocks resulting in distinct vulnerabilities (Taylor 2014). Societal relationships enable and limit efforts (Toulmin 2011). For example, commercialising requires farmers to have the skills to make use of market information and climate information, which is unevenly available to men and women, as is the social capital to act upon the information.

In Zambia and Ghana, van Koppen et al. found that men could more easily adopt irrigation technologies because they had lighter workloads and easier access to equipment, inputs, finance, public support, transportation and markets (2012, in Bernier et al. 2015). Similarly, Magnan et al. found that poor women in India had larger networks, but men had more effective and smaller networks that connected them to knowledge about new agricultural technologies (2013, in Bernier et al. 2015).

Some of these themes are echoed in the literature on gender and climate-smart agriculture, for example in suggestions that gender relations is a key factor in

(18)

determining the feasibility of climate-smart agricultural practices (Thornton et al. 2017). One example is Conservation Agriculture (CA), widely considered an attractive climate-smart agriculture practice (Thierfelder et al. 2017). CA has been promoted across East and Southern Africa, while also being critiqued for disadvantaging women and children due to its higher labour requirements for land preparation and weeding. However, while the literature demonstrates that the gender barriers result in uneven commercialisation and uneven climate change impacts, it also points to the possibility that because women have not entirely entered the commercial approach they may be spared the negative implications of climate change on single crop, market-based approaches.

Kirsten et al. (2013) identify the socioeconomic characteristics of smallholder producers that are determinants for success in commercialisation. Their discussion includes how gender influences the types of crops grown (less marketable varieties) amidst other variables including isolation from markets, household asset base, and access to agricultural support services (ibid.). While this discussion is helpful to understanding implications for livelihood resilience along commercialising pathways, further research is needed to understand the uneven processes of commercialisation and the differentiated implications of climate change.

1.6 Climate change impacts on

commercialisation from the value

chain perspective

Agricultural value chains encompass the flow of products and services including information between farmers and consumers. The agricultural value chain remains a core element of agricultural commercialisation as it defines the various stages through which agricultural products pass to achieve specified commercial value (Lawton 2011). Value chains enable farmers – whether small or large scale – to gain added value on their products, thus enhancing the market reach and potentially, depending on the commercialisation activities chosen, the resilience of products. However, climate change impacts are largely embedded within value chain stages including production, processing and consumption. As illustrated in Figure 2, the three value chain levels are connected to each other through flow of materials and products and this depicts the fact that climate change impacts are also linked across the

value chain. Studies have therefore emphasised that looking into the climate change impacts on agricultural commercialisation through the value chain lens enables farmers to analyse vulnerability to climate, identify hotspots for risk across the whole chain, and identify opportunities for new products and markets for a more resilient commercialisation pursuit.

Primary production represents the first stages of a value chain where agricultural products are produced through natural bio-geochemical processes. This involves intense interactions of ecosystem services including soil formation, nutrient cycling (for crops) and biological processes for livestock to produce agricultural products. As already discussed, there is a general consensus within a wide range of literature (e.g. IPCC 2014; Malhi and Wright 2004; UNDP 2007) that primary production of crops and livestock products is already suffering the most impacts of climate change. Climate change directly affects critical ecosystems services such as rainfall availability, nutrient cycling, etc. which in turn results in decline in yields. Varying rainfall over time and space already causes severe declines in yields of rainfed crops resulting in hunger among smallholder farmers who are the main food producers and the majority in Africa (World Bank 2008). Further incidences of extreme events such as heat waves and pests and diseases further pose threats to crop yields. The processing and value addition stage of commercialisation is largely characterised by the refining of original agricultural products – for example maize cobs, tea leaves, coffee beans (for crops) or milk, meat (for livestock) – into commodities with added value, such as enhanced durability, packaging, nutrition, etc. Processing takes place either through advanced equipment or machinery, particularly for large-scale farmers, while for small-scale farmers, basic equipment and processes are employed. Climate change impacts this level indirectly, to the extent that it disrupts the supply chains both in terms of reducing raw materials and infrastructural efficiency. For instance, most agricultural industries in Africa utilising water now experience increased costs of production, for example water constraints especially. Further, several African countries are now experiencing climatically induced human diseases, health and safety hazards for the agricultural workforce, and reduced timeliness and efficiency in commercialisation processes (UNDP 2007). The consumption level of commercialisation occurs at the end stage where processed commodities are

(19)

utilised by consumers in various ways, either for direct consumption or other utilities. At this level, climate impacts on commercialisation are indirect – mainly manifesting via impacts on consumer/households’ access to commodities as well as impacts on the wellbeing and lifestyle of consumers which affects consumption patterns and commodity market value. In many African countries, climate change has reportedly disrupted agricultural commodity distribution and networks, affecting delivery times and causing production interruptions, poor commodity access and sales losses. Further, climatically vulnerable communities are often forced to change their lifestyles as part of adaptation, including eating habits that shifts focus from certain commodities. According to Amado and Adams (2011), communities provide the ‘social licence’ for businesses to operate and so their vulnerability to climate change posits commercialisation risks.

Despite the novelty of the value chain approach to understanding climate impacts on agricultural commercialisation, a number of challenges exist. The key ones include poor access to climate-related data for specific value chain levels, especially in Africa, and uncertainties associated with future climatic changes

makes it difficult to predict the actual costs of climatic uncertainties. Nonetheless, we think that the value chain approach will be a critical tool to utilise for an in-depth analysis of climate change impacts on agricultural commercialisation under the APRA project.

1.7 How resilient to climate change is

modern agriculture?

Farmers pursuing any of the four pathways to commercialisation (see Box 1) may employ ‘modern’ techniques such as using improved seed varieties, non-organic fertilisers, and mechanised farming methods. Alternatively, farmers may intensify production for surplus sale using traditional seed varieties, conservation agriculture techniques, and other sustainable intensification methods, such as permaculture or organic farming. Techniques that improve water retention and reduce run-off, such as integrated soil fertility management practices or soil and water conservation, can stabilise yields whether inputs are organic or modern.

Still, there seems to be a divide in the literature between those who either argue or assume that pursuing modern techniques is better for farmers, and conversely, those

Figure 2: Schematic illustration of climate change interactions with agricultural value chain

Primary production Processing Consumption Raw m ate ria ls Agric . i np ut s Ag ric . c om m od iti es W ast es Value chain Direct impact: ecosystem

services

Indirect impacts: supply chain and prices

Climate impacts

(20)

who make the argument that modern techniques are more climate-sensitive and require optimum conditions, and are more implicated in driving anthropogenic climate change the more industrial the scale at which they are practised. Particularly in Africa, the assumption often leans toward pro-modernisation in part because yield levels are much lower than the rest of the world, so food production can grow to a higher potential before reaching biophysical limits (Grassini et al. 2013; Jayne et al. 2016; Livingston et al. 2011). However, much of the climate change literature argues that while heavy use of inputs may help farmers to offset some of the effects of climate change in the near term through higher incomes, in the medium-to-long term, it is an unsustainable approach to climate adaptation. More diverse farms, which intercrop, manage soils with organic inputs, maintain seed diversity and keep indigenous livestock breeds, may be more drought, flood and pest resilient in the longer term.

Fortier and Trang make the argument that a reliance on inputs, particularly agrochemicals and modern rice varieties in the case of Vietnam, impoverishes environmental and social ecosystems:

We are argue that this modernisation has locked both family and large-scale farms into technological path dependencies of energy- and input-intensive production, notably for agrochemicals, biotechnologies and water. This in turn, has led to a vicious circle of induced systemic fragility through engineered landscapes, reduced agro-biodiversity, and weakened social networks, knowledge and skills. As a result, Vietnam has become more sensitive to structural changes and less able to adapt to the unpredictable context of climate instability. Beyond those systemic contradictions, however, we argue that the most damaging impact of modernisation under Doi moi (market reform) has been to generate a new class dynamic and transform state-society relations in ways that now undermine the country’s ability to respond to climate change (Fortier and Trang 2013: 2).

The authors argue that Vietnam’s agricultural modernisation has ‘made the country increasingly reliant on complex processes and has locked in various technological path dependencies’ while reducing agro-biodiversity and weakening social networks, knowledge and skills (ibid.: 1). Likewise, Sokoni (2008) makes the argument that commercialisation aggravates the erosion of the natural resource base.

One of the fundamental components of commercialisation, access to irrigation, could be seen

as an important aid to adaptation, to the extent that it reduces smallholder dependency on the vagaries of rainfall and shields them from the sometimes-devastating impacts of a failed harvest. However, the use of irrigation can also in some contexts heighten the vulnerability of smallholder farmers to precisely the kinds of environmental phenomena, such as droughts, that climate change is projected to exacerbate across many parts of Southern Africa. The experience of farmers in Andhra Pradesh provides a cautionary tale in this regard, as the work of Marcus Taylor (2013, 2014) demonstrates. Against a background of liberalisation policies from the 1990s onwards, farmers in Andhra Pradesh have responded by planting commercial and non-food grain crops such as paddy, groundnut, oil seeds, vegetables and cotton. This resulted in a much greater use of high-yield varieties of seed in combination with irrigation and inputs such as pesticides and fertilisers, to the extent where farmers were spending up to 35 percent of their income on such inputs. Often, loans to pay for these inputs would be taken out from the very people to whom farmers were looking to sell their crops, and interest rates charged would exceed the profits made from the sale in some cases. Compounding this dependency, in 2012–13, was the presence of severe drought in Andhra Pradesh, which brought about water shortages that meant insufficient water was available not only for irrigation purposes, but also for livestock. In some cases, this led to distress sales of cattle at very low prices that Hyderabad merchants subsequently profited from substantially. This was also a case of relational vulnerability, where the profits that merchants made, which consolidated a livelihood that was not very sensitive to climate impacts in itself, came at the expense of resilience to climate impacts of the farmers involved in distress sales. The cycle of poverty for these farmers is thus renewed with the stripping of their assets.

The implication here is that commercialisation pathways which do not take sufficiently into account the agro-environmental conditions in which they are being applied may end up contributing to heightened vulnerability to climate impacts, and that commercialisation activities may simultaneously strengthen the resilience of some directly at the expense of others. It would be useful in work streams 1 and 2 to investigate the prospects for such uses in the case study countries that APRA will focus on. It will also be important to understand, in the context of work stream 3, the extent to which the policy environment is likely to favour or dismiss commercialisation strategies which are more responsive to agro-environmental considerations.

(21)

Brooks et al. argue that modern agriculture is based on a ‘concept of progress and the belief that societies tend to become more advanced through endogenous processes of social change leading to an increasing separation of humans from nature’ (2009: 744). They argue that development policies supporting modernising agriculture have ‘exacerbated social and/ or environmental vulnerability to previous or potential future changes in climate’ (ibid.: 745). They contend also that development sees climate concerns as a threat, with the potential to undermine development, instead of development being designed around environmental constraints and opportunities. ‘This may require people to choose not to maximise production in the short term… [but] may prevent societies (from) becoming dependent on levels of production that require “optimum” climatic conditions and which are not suitable in the medium to long term’ (ibid.: 755). Brooks et al. (2009) provide three modernisation examples of areas where policies have promoted a shift from subsistence to commercial agriculture with negative implications in light of climate change. First, the authors discuss the Sahel where policies to modernise traditional livelihoods by expanding agriculture, reducing pastoralism, and moving away from subsistence to commercialisation have resulted in livelihood approaches that are less suited to changing climate conditions. Second, the authors discuss Brazil, where ‘military-initiated resettlement and large investment in infrastructure from the late 1960s to 1980s led to unsustainable and inequitable patterns of development that are still continuing today’ (ibid.: 745). Third, they discuss the Kenyan government’s Kenya Vision 2030, which envisions a modern development of commercialised agriculture into new land and semi-arid areas. The authors argue that these approaches to commercialisation and modernisation fail ‘to consider or plan for variations in climatic and environmental conditions on timescales longer than those associated with seasonal or inter-annual variability’, and are, thereby, not appropriate for a world in which the climate is changing (ibid.: 746).

Ensor (2009) argues that modern, industrial agro-ecosystems have weak resilience and biodiverse agriculture has stronger resilience. He argues that agro-ecological approaches that build resilience include: traditional complex ecosystems and diversified cropping; use of local genetic diversity; enhanced soil organic matter; intercropping; agroforestry and mulching; and home gardening. Ensor’s argument hinges on the concept that ‘traditional agro-ecosystems

are less vulnerable to catastrophic loss because they grow a wide range of crops and varieties in various spatial and temporal arrangements’; agro-ecological methods better withstand shocks and stresses through diverse practices that enrich resources (ibid.: 12). For example, Ensor (2009) cites a project in the Sahel that focused on improving soil quality, integrating stall-fed livestock into crop systems to improve the use of manures and phosphates, adding legumes and green manures, incorporating water harvesting systems, and developing effective composting systems, reporting results at a 75–195 percent improvement in millet and groundnut yields. According to programme participants, yields are variable according to drought, but poor rains have a negative effect instead of total crop failure. He furthers his case with evidence from Central America in the wake of Hurricane Mitch when farmers who practiced agro-ecology suffered fewer economic losses; their land was found to have 20–40 percent more topsoil, which had greater capacity to retain large quantities of moisture and resulted in less erosion (ibid.).

Ensor (2009) further cites two long-running studies that compared organic and conventional farms, which found higher levels of organic matter in the organically farmed soils, which one study found to be of consequential benefit during drought years. His review includes discussion of studies comparing organic and conventional farm systems, as well as studies comparing diversified versus single crop farms, arguing that yield increases can and do happen in diversified organic farms in a more sustainable manner. His argument concludes that in addition to increasing yields, focusing efforts on organic matter accumulation and natural control mechanisms, versus inputs, conserves resources and allows for regeneration and higher adaptive capacity to shocks and stresses (ibid.). If commentators, such as Jonathan Ensor, who suggest that a focus on organic farming is likely to increase resilience in the face of climate impacts are proved right, then the implications for pathways of commercialisation in sub-Saharan Africa are difficult to reconcile. On the one hand, purely from an adaptation perspective, it is tempting to recommend organic farming methods. However, there is not universal agreement that these methods deliver better yields than modernised agricultural methods relying on inputs such as fertilisers and modified seeds. Advocating farming methods which offer lower yields per hectare is not automatically compatible with commercialisation